Educational technologies, such as teacher dashboards, are being developed to support teachers' instruction and students learning. Specifically, dashboards support teachers in providing the just-in-time instruction needed by students in complex contexts such as science inquiry. In this study, we used the Inq-Blotterteacher-alerting dashboard to investigate whether teacher support elicited by the technology influenced students' inquiry performance in a science intelligent tutoring system, Inq-ITS. Results indicated that students' inquiry improved after receiving teachers' help, elicited by the Inq-Blotteralerts. This inquiry improvement was significantly greater than for matched students who did not receive help from the teacher in response to alerts. Epistemic network analyses were then used to investigate the patterns in the discursive supports provided to students by teachers. These analyses revealed significant differences in the types of support that fostered (versus did not foster) student improvement; differences across teachers were also found. Overall, this study used innovative tools and analyses to understand how teachers use this technological genre of alerting dashboards to dynamically support students in science inquiry. [This is the online version of an article published in "Journal of Learning Analytics (ISSN 1929-7750)."]